Knowledge
Everything Kapa answers starts in this section: you connect your knowledge sources, and ingestion builds them into one knowledge base and keeps it current as they change. Each connector is optimized for consumption by AI, with automatic refreshes and configurable filters (for example, masking PII in internal sources). Everything connected here becomes searchable by the engine, and every answer cites it.
Searching the knowledge base is the other half of the engine, agentic retrieval, one section down. This section covers what goes in and how it stays healthy.
Connecting your first source takes a few minutes: follow Index your first source, then pick further connectors from the catalog.
Once sources are connected, ingestion keeps the index correct and current automatically; how data ingestion works explains the pipeline, and refreshes lists the per-source schedules. You manage sources from the Sources view (permissions and editing), and when one project serves several products or versions, you organize them with source groups. Answer quality starts with what you index; see writing documentation for AI.